Substitutability refers to the extent to which one alternative recreat
ion activity can effectively replace another in terms of its ability t
o produce desired experiences and benefits (e.g., recreation satisfact
ion). Past research into substitution has generally focused on develop
ing typologies of substituable activities or on satisfactions obtained
and needs fulfilled by activities. These have not, however, proven ad
equate in explaining substitution decisions. To better understand recr
eation substitution, we propose an approach that considers and evaluat
es several elements of recreation choice and their attributes in combi
nation. Specifically, we examine substitutability in terms of differen
ces in activity choice options, social group with which the preferred
activity was to have been pursued, and level of specialization in the
initially preferred activity. Data for this study were collected by us
ing a mail survey (with a scenario) sent to a sample of recreational a
nglers. The instrument was designed as a 4 (specialization levels) x 4
(social groups) x 4 (alternative activities) factorial, which permitt
ed identification of main and interaction effects due to these three i
ndependent variables. The dependent variable was the probability of su
bstituting an alternative activity in place of the planned activity. R
esults showed significant main effects for the social group and activi
ty treatments, but no significant main effect for the specialization t
reatment. Significant interaction effects were found between social gr
oup and alternative activity, between social group and specialization
level, and between alternative activity and specialization level. Impl
ications for future research are discussed, as are a number of issues
concerning recreation substitution.